US20120146802A1 - Methods for Forewarning of Critical Condition Changes in Monitoring Civil Structures - Google Patents
Methods for Forewarning of Critical Condition Changes in Monitoring Civil Structures Download PDFInfo
- Publication number
- US20120146802A1 US20120146802A1 US12/962,787 US96278710A US2012146802A1 US 20120146802 A1 US20120146802 A1 US 20120146802A1 US 96278710 A US96278710 A US 96278710A US 2012146802 A1 US2012146802 A1 US 2012146802A1
- Authority
- US
- United States
- Prior art keywords
- forewarning
- data
- signal
- mechanical power
- accelerometer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
- G01M5/0033—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings by determining damage, crack or wear
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
- G01M5/0008—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings of bridges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
- G01M5/0066—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings by exciting or detecting vibration or acceleration
Definitions
- the field of the invention is computer methods for monitoring and forewarning of condition changes, including critical events, such as structural failures in structures such as bridges, buildings, dams, tunnels and other civil structures.
- the technology can also be applied to large manufactured items such as airplanes, and mining and construction equipment.
- Welch et al. U.S. Pat. No. 6,460,012, disclosed an apparatus for the detection of an indicator of imminent failure due to crack growth in structural elements.
- the apparatus included sensors for sensing physical data factors, a processor for computing a relationship between the physical data factors and an indicator of the failure event.
- U.S. Pat. No. 6,012,337 discloses physical integrity monitoring of bridges and buildings using optical sensors to sense directional movements of the structures in a plurality of axes.
- Strasser et al. U.S. Pat. No. 6,292,108, discloses a data acquisition system for collecting data from a plurality of sensor units for sensing damage in civil structures resulting from earthquakes, for example. Strasser et al. discloses the use of accelerometers as sensors for collecting mechanical vibration data. Di Marzio, U.S. Pat. Pub. No. US2004/0078170 discloses a system with sensors disposed on a structure, a central repository and a mesh network for communicating the data from the sensors to the central repository.
- the present invention improves over the prior art by providing an improved method of processing sensor data collected from sensors on a physical structure to provide a forewarning of a critical event signaling deterioration or failure of the structure.
- the present invention provides a failure forewarning by acquiring accelerometer data, which is then converted into mechanical vibration power.
- the latter is to be distinguished from mechanical vibration data of the type utilized in the prior art.
- P mechanical power
- failures can be due to various causes, such as low-temperature creep crack growth, stress corrosion cracking, and fatigue cracking in various materials (aluminum, steel, and glass-fiber/polymer matrix composite); single and multiple-site or widespread fatigue damage; and constant amplitude loading, as well as periodic overloads.
- the method can be further implemented with various network configurations known in the art.
- the acceleration data can be transmitted through a local network and then through a wide area network to a host computer for further processing to compute and provide forewarning at a central location.
- Many other types of networks including wired and wireless Internet networks, cellular, satellite and cable networks and other networks to be developed in the future can also be utilized.
- the method can also be practiced within a single sensor module attached to the civil structure, with forewarning signals transmitted to an appropriate location either nearby or through a network to a central monitoring location.
- FIG. 1 is a schematic diagram of a network for collecting sensor data from a physical structure
- FIG. 2 is a block diagram of one of the sensors seen in FIG. 1 ;
- FIG. 3 is a block diagram of a central computer system for collecting data from the sensors through the network, and for providing a forewarning signal;
- FIG. 4 is a flow chart of the processing of the sensor data at the central computer system to provide the forewarning signal
- FIG. 5 is a graph of hysteresis strain energy as a function of fatigue cycles.
- FIG. 1 illustrates a civil structure, in this example a bridge 10 .
- Sensor modules 12 are mounted on the bridge members 11 by means for mounting 19 such as brackets, fasteners, adhesives, frames and panel enclosures, and others of a type known in the art.
- the sensors communicate by a local area network 13 , preferably a wireless network, but a wired network is also possible, to a gateway radio receiver 14 , which may also transmit signals to the sensor modules 12 in a two-way network.
- the range of transmission is limited by standards governing the particular type of local area wireless network and is preferably 100 meters or less, but may extend up to 400 meters or more in some other embodiments.
- the gateway 14 is connected to a router 15 for accessing a wide area network 16 , such as the Internet. This allows for communications of data to a host computer system 17 at a central location with personnel for monitoring conditions at one or more civil structures 10 .
- the sensor module 12 includes a multi-axis, typically a three-axis accelerometer 20 , which is mounted to the civil structure 10 so as to sense motion along each of the axes and to generate axis acceleration data.
- a multi-axis typically a three-axis accelerometer 20
- Many types of accelerometers are known in the art, including those available from Analog Devices and others disclosed in U.S. Pat. No. 6,292,108, cited above.
- the sensor module 12 includes a microelectronic central processing unit (CPU) 21 , also referred to more generally and more commonly as “a processor.”
- the CPU 21 is interfaced to the accelerometer 20 through a suitable I/O interface 22 according to specifications of the CPU 21 and the accelerometer 20 .
- the CPU 21 is also connected by suitable buses to a radio transmitter/receiver circuit 23 of a type known in the art for converting between digital data and radio signals encoded with the data.
- the radio transmitter/receiver circuit 23 transmits and receives these signals through a suitable antenna 24 .
- the CPU 21 is operated by executing instructions in a control program 25 stored in memory 26 .
- the memory 26 is shown generally and may include both memory on-board, and external memory in addition to that available on-board the CPU, and also includes memory for data storage in addition to the control program.
- the sensor module 12 also has a power source 27 , which may be a battery, a connection to an AC power outlet or a solar power transducer and storage circuit.
- the sensor module 12 also includes a small visual display 28 and an alarm 29 having an audible mode of operation and an inaudible, vibratory mode of operation.
- the host computer system 17 which can be addressed at an Internet address, has a network server portion 30 for transmitting and receiving data on the Internet.
- the host computer system also has a microelectronic central processing unit (CPU) 31 , also referred to more generally and more commonly as “a processor.”
- the CPU 31 is operated by executing instructions in a control program 32 stored in memory 33 .
- the memory 32 is shown generally and may include both memory on-board, and external memory in addition to that available on the CPU 31 , and also includes memory for data storage in addition to the control program.
- the host computer system 17 includes a visual display 34 and also has the ability to emit audible sounds through a built-in speaker 36 .
- the host computer system 17 may also include a database 35 , which can be stored in a computer memory or in an external bulk storage memory device of a type known in the art.
- the methods disclosed herein can be used to continuously monitor civil structures or large manufacture items to detect impending failures, thereby detecting problems before they occur and thus allowing condition-based maintenance.
- the method is based on a premise that hysteresis strain energy (HSE) per unit time is the same as mechanical power invested in crack growth per unit time.
- HSE hysteresis strain energy
- U.S. Pat. No. 6,460,012 assigned to the assignee herein, discloses a method for analysis of a load, L(t), and deflection, ⁇ (t) versus time, t.
- the deflection can be expressed in relation to strain energy, U, by integration over the N th load-unload cycle ( ⁇ C ) of the structure as follows:
- the Griffith-energy criterion states that a crack cannot grow unless the strain energy per unit of crack length, dU/da, equals (or exceeds) the Griffith energy:
- Equation (3) is a material property, and the subscript, 1, is for mode-one failure (under tension). Substitution of Equation (4) for the case of equality into Equation (3) yields:
- Each load-unload cycle spans a cycle time, ⁇ t, and corresponds to multiplication of the left-hand side of Eq. (5) by dN/dt, while the right-hand side is divided by ⁇ t.
- the resultant form is then:
- Eq. (9) means that the conversion of tri-axial acceleration (vector) to mechanical power (a scalar quantity) allows direct monitoring of the crack growth rate, as (for example) illustrated in FIG. 5 . Because mechanical power (P) is a scalar quantity, this result is independent of the orientation of any tri-axial accelerometer in the case where acceleration is sensed along three axes which are orthogonal to each other.
- FIG. 5 shows that structural failure is indicated by a statistically significant change in the hysteresis strain energy (HSE), as an outlier in HSE slope beyond four standard deviations ( 4 ⁇ ).
- HSE hysteresis strain energy
- FIG. 5 shows the slope of HSE versus number of fatigue cycles for corroded (10% thickness reduction), un-notched aluminum sample with typical failure forewarning in the last 1-5% of life. More specifically, the top curve 61 in FIG. 5 is +4 ⁇ above the running value of the slope of HSE. The bottom curve 62 in FIG. 5 is ⁇ 4 ⁇ below the running value of the slope of HSE. The middle curve 63 is the HSE slope, which falls abruptly below the ⁇ 4 ⁇ level at 49 000 cycles, as an indicator of structural failure.
- HSE hysteresis strain energy
- FIG. 4 illustrates a method for providing forewarning of a critical event, such as a structural failure in a civil structure. This method is performed by one or more processors in executing instructions in one or more control programs, depending on the network configuration.
- the acceleration data is transmitted from the sensors 11 to the host computer system 17 which performs the operations diagrammed in FIG. 4 .
- the operations are performed by each of the sensors 11 and a forewarning signal is signaled from the sensors 11 in the first instance, and may be relayed to the host computer.
- artifacts are removed from the data with a zero-phase quadratic filter that performs better than conventional filters.
- This filter uses a moving window of 2w+1 points of data, with the same number of data points, w, on either side of a central point.
- a parabola is fitted in the least-squares sense to these data points, and the central point is selected to estimate the low-frequency artifact, f i .
- the residual (artifact-filtered) signal, g i e i ⁇ f i , has essentially no low-frequency artifact activity. All subsequent analysis uses this artifact-filtered data, g i . This is known in the art from U.S. Pat. No. 5,626,145.
- decision block 43 a quality check is made of the entire dataset as described in U.S. Pat. No. 5,815,412, col. 12, line 64 to col. 14, line 9. If poor data quality is detected the dataset is discarded, as represented by the “FAIL” result branch and the process returns to obtain a new set of data at process block 41 . If the quality is suitable as represented by the “PASS” result branch, the routine proceeds to the next decision block 44 .
- Each artifact-filtered point is converted into a discrete symbol, s i , as one of S different integers in the range, 0 ⁇ s i ⁇ S ⁇ 1.
- s i INT[S (g i ⁇ g min )/(g max ⁇ g min )] for g i ⁇ g max
- equiprobable symbols are formed by ordering all N of the base case artifact-filtered, time-serial data points from the smallest to largest value.
- the first N/S of these ordered values correspond to the first symbol, 0.
- Ordered data values (N/S)+1 through 2N/S correspond to the second symbol, 1, and so on.
- Equiprobable symbols have non-uniform partitions in the signal amplitude with the same occurrence frequency of g i values by construction, and thus have no information about the dynamical structure.
- time-serial s i -data are converted into a geometric object via phase-space (PS) reconstruction that uses time-delay vectors with the form,
- the multi-channel PS vector takes the form:
- y ( i ) [ s i (1), s i+ ⁇ (1), . . . , s i+(d ⁇ 1) ⁇ (1), . . . , s i ( Ch .), s i+ ⁇ ( Ch .), . . . , s i+(d ⁇ 1) ⁇ ( Ch .)] (11)
- the symbol, s(k) denotes values from the k-th channel, i.e., 1 ⁇ k ⁇ Ch., where Ch. is the total number of data channels.
- the choice of lag, ⁇ , and embedding dimension, d determines how well the PS reconstruction unfolds the dynamics. An excessively large embedding dimension could result in over-fitting of real data with finite length and noise.
- Conversion of the time-serial data into discrete PS states allows the construction of a statistical distribution function (DF), as represented by process block 48 , by counting the number of PS points that occur in each bin.
- the population of the ith bin of the DF is denoted as Q i , for the base case, and R i for a test case, respectively.
- An (un)changing DF indicates (un)altered dynamics.
- a check is made to determine if the baseline number “B” of cutsets have been processed, as represented by decision block 49 .
- the dissimilarity functions, DFs, from first B cutsets are saved as represented by blocks 50 in FIG. 1 , to represent the nominal baseline dynamics.
- decision block 51 when the number of cutsets equals the statistical sample number, B, the baseline dissimilarity calculations are completed as represented by process block 52 .
- the baseline DFs correspond to the nominal-state dynamics. These B baseline DFs are exhaustively compared to one another in pair-wise fashion. This comparison is via the dissimilarity measures (DM) of Eqs. (12)-(15), as represented by block 54 .
- V is the mean value of the dissimilarity measure, which is computed by comparison among the B(B ⁇ 1)/2 unique combinations of the B base case cutsets, along with a corresponding standard deviation, ⁇ 1 , for each DM from the set, V.
- the dissimilarity measures are computed as represented by block 53 , between DFs for the baseline, Q i , and test case, R i , respectively.
- One set of dissimilarity measures (DM) for the non-connected phase space is:
- the summations in Eqs. (12)-(13) run over all of the populated PS bins.
- the ⁇ 2 statistic is one of the most powerful, robust, and widely used tests for dissimilarity between two distribution functions (DFs).
- ⁇ 2 is not an unbiased statistic for accepting or rejecting a null statistical hypothesis but rather is a relative measure of dissimilarity between the two distribution functions (DFs).
- the L 1 distance is the natural metric for the distribution functions (DFs) by its direct relation to the total invariant measure on the attractor. These measures account for changes in the geometry and visitation frequency of the attractor. Consistent calculation requires the same number of points in both the base and test case distribution functions (DFs), identically sampled; otherwise the distribution functions must be resealed.
- Y(i) is a 2d-dimensional, connected-phase-space (CPS) vector.
- Q and R denote the CPS DFs for the base case and test case, respectively.
- phase space dissimilarity measures (PSDM). Their definitions allow PSDM to detect transitions between regular and chaotic regimes, as well as to discriminate between different chaotic regimes.
- the dissimilarity measures are normalized by comparing each of the B baseline cutsets to each (ith) test case cutset, and then computing the corresponding average dissimilarity value, V i , of the ith cutset as represented by process block 55 .
- Each contiguous, non-overlapping test case cutset is compared to each of the B base case cutsets, to obtain the corresponding average dissimilarity value, V i , of the i-th analysis window for each dissimilarity measure.
- a statistically significant trend in the normalized dissimilarity measure indicates equipment degradation for failure forewarning.
- a forewarning threshold is then developed by computing a cumulative sum of the composite measure of dissimilarity ( ⁇ i C i ) over a series of time frames.
- a straight-line fit is applied to the cumulative sum of composite measures ( ⁇ i C i ) using a least squares calculation. Then, a standard deviation ( ⁇ 2 ) is computed between the straight-line fit and the cumulative sum ( ⁇ i C i ).
- Event forewarning corresponds to the difference between the cumulative sum ( ⁇ i C i ) and the straight line fit exceeding a forewarning threshold by more than a selected number of standard deviations.
- a forewarning signal of an impending critical event is provided to a human observer, as represented by I/O block 58 , through one of or a combination of a visual signal, an audible signal or a tangible signal.
- An end-of-life failure forewarning threshold can also be determined and an end-of-life failure forewarning signal issued based on a number of successive occurrences of C i above a failure threshold over a selected time period, with the failure threshold being computed as disclosed in U.S. Pat. No. 7,209,861, col. 10, line 25 to col. 11, line 26.
- the failure threshold being computed as disclosed in U.S. Pat. No. 7,209,861, col. 10, line 25 to col. 11, line 26.
- a failure forewarning threshold for detecting failure forewarnings corresponds to a minimum in said standard deviation ( ⁇ 2 ) in relation to said straight-line fit of ⁇ i C i , which is the cumulative sum of the composite measure of dissimilarity over a sampling period.
- a failure value threshold is determined based on a number of failure forewarnings occurring after this minimum in the standard deviation is detected.
- a forewarning of the impending critical event is signaled to a human observer, as represented in block 58 , after either one of 1) a selected number of occurrences of the composite measure of dissimilarity (C i ) above the forewarning threshold or 2) after a selected number of occurrences of the composite measure of dissimilarity (C i ) above the failure threshold.
- decision block 59 if the result from executing decision block 57 is “No,” a check is then made for more data, and if more data is available, the routine returns to block 41 , to continue the analysis for the next dataset via the above algorithm, and then combine the results for all of the datasets. If no more data is available, the execution is ended as represented by block 60 .
- This method can be applied to any large physical structure (e.g., bridge, aging aircraft, cranes, spacecraft, buildings) to forewarn of an impending structural failure under arbitrary loading, thus reducing the cost of inspections and repairs and conserving critical infrastructure.
- the method to perform this analysis has been demonstrated on desktop computers.
- Modern hand-held devices e.g., PalmPRE and iPhone
- Such hand-held devices have embedded GPS (to provide geo-location), wireless communication (to send the forewarning information), and a tri-axial accelerometer (to acquire the acceleration data). Therefore, these devices could also be re-packaged and utilized as the sensors 11 .
- a cellular network can be used in place of the Internet wide area network disclosed herein between the sensors 11 and a host computer in a cellular network.
- the CPU 21 or the CPU 31 can transmit a signal through an output interface circuit of a type well known in the art to alarm devices 29 or 36 , respectively, to emit an audible signal, or a tangible vibration signal, or to a screen display 28 , 34 to provide visual signals or visual data.
- the output devices provide forewarnings prior to a critical event or prior to a failure event as described above.
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
Description
- This invention was made with assistance under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The Government has certain rights in this invention.
- The field of the invention is computer methods for monitoring and forewarning of condition changes, including critical events, such as structural failures in structures such as bridges, buildings, dams, tunnels and other civil structures. The technology can also be applied to large manufactured items such as airplanes, and mining and construction equipment.
- Hively et al., U.S. Pat. No. 5,815,413, disclosed methods of forewarning in medical patients, and also described the applicability of nonlinear techniques to monitor machine conditions using accelerometer data. Hively et al., U.S. Pat. No. 6,484,132, provided improvements in these methods of forewarning. Hively, U.S. Pat. No. 7,139,677, introduced a composite measure of dissimilarity (C). This composite measure of condition change (C) was calculated from the sum of the four normalized measures of dissimilarity, including U(χC 2) and U(LC) from the connected phase space and including U(χN 2) and U(LN) from the non-connected phase space. This was developed further across multiple data channels in Hively, U.S. Pat. No. 7,209,861, where the composite measure of dissimilarity, (C), was used to provide an end-of-life forewarning factor (G).
- Welch et al., U.S. Pat. No. 6,460,012, disclosed an apparatus for the detection of an indicator of imminent failure due to crack growth in structural elements. The apparatus included sensors for sensing physical data factors, a processor for computing a relationship between the physical data factors and an indicator of the failure event.
- Hodge, U.S. Pat. No. 6,012,337 discloses physical integrity monitoring of bridges and buildings using optical sensors to sense directional movements of the structures in a plurality of axes.
- Strasser et al., U.S. Pat. No. 6,292,108, discloses a data acquisition system for collecting data from a plurality of sensor units for sensing damage in civil structures resulting from earthquakes, for example. Strasser et al. discloses the use of accelerometers as sensors for collecting mechanical vibration data. Di Marzio, U.S. Pat. Pub. No. US2004/0078170 discloses a system with sensors disposed on a structure, a central repository and a mesh network for communicating the data from the sensors to the central repository.
- The present invention improves over the prior art by providing an improved method of processing sensor data collected from sensors on a physical structure to provide a forewarning of a critical event signaling deterioration or failure of the structure.
- The present invention provides a failure forewarning by acquiring accelerometer data, which is then converted into mechanical vibration power. The latter is to be distinguished from mechanical vibration data of the type utilized in the prior art.
- Mechanical power data can be obtained by conversion of tri-axial acceleration data, a three-dimensional vector quantity, to mechanical power data, P, a scalar quantity by the dot product expression of two vectors, P=A·∫dt A, where P is mechanical power, and A is an acceleration vector quantity. By using one or more sensors utilizing tri-axial accelerometers on the structure, and collecting sensor data, mechanical power (P) data can be obtained from processing the sensor data. Because this is a scalar quantity, this result is independent of the orientation of any tri-axial accelerometer providing the data. This data can be further processed to provide a forewarning of critical events including impending structural failures.
- These failures can be due to various causes, such as low-temperature creep crack growth, stress corrosion cracking, and fatigue cracking in various materials (aluminum, steel, and glass-fiber/polymer matrix composite); single and multiple-site or widespread fatigue damage; and constant amplitude loading, as well as periodic overloads.
- The method can be further implemented with various network configurations known in the art. The acceleration data can be transmitted through a local network and then through a wide area network to a host computer for further processing to compute and provide forewarning at a central location. Many other types of networks including wired and wireless Internet networks, cellular, satellite and cable networks and other networks to be developed in the future can also be utilized.
- The method can also be practiced within a single sensor module attached to the civil structure, with forewarning signals transmitted to an appropriate location either nearby or through a network to a central monitoring location.
- Other objects and advantages of the invention, besides those discussed above, will be apparent to those of ordinary skill in the art from the description of the preferred embodiments, which follows. In the description reference is made to the accompanying drawings, which form a part hereof, and which illustrate examples of the invention. Such examples, however, are not exhaustive of the various embodiments of the invention, and therefore reference is made to the claims, which follow the description for determining the scope of the invention.
-
FIG. 1 is a schematic diagram of a network for collecting sensor data from a physical structure; -
FIG. 2 is a block diagram of one of the sensors seen inFIG. 1 ; -
FIG. 3 is a block diagram of a central computer system for collecting data from the sensors through the network, and for providing a forewarning signal; -
FIG. 4 is a flow chart of the processing of the sensor data at the central computer system to provide the forewarning signal; and -
FIG. 5 is a graph of hysteresis strain energy as a function of fatigue cycles. -
FIG. 1 illustrates a civil structure, in this example abridge 10.Sensor modules 12 are mounted on thebridge members 11 by means for mounting 19 such as brackets, fasteners, adhesives, frames and panel enclosures, and others of a type known in the art. The sensors communicate by alocal area network 13, preferably a wireless network, but a wired network is also possible, to agateway radio receiver 14, which may also transmit signals to thesensor modules 12 in a two-way network. The range of transmission is limited by standards governing the particular type of local area wireless network and is preferably 100 meters or less, but may extend up to 400 meters or more in some other embodiments. Thegateway 14 is connected to arouter 15 for accessing awide area network 16, such as the Internet. This allows for communications of data to ahost computer system 17 at a central location with personnel for monitoring conditions at one or morecivil structures 10. - Referring next to
FIG. 2 , one of thesensor modules 12 is illustrated. Thesensor module 12 includes a multi-axis, typically a three-axis accelerometer 20, which is mounted to thecivil structure 10 so as to sense motion along each of the axes and to generate axis acceleration data. Many types of accelerometers are known in the art, including those available from Analog Devices and others disclosed in U.S. Pat. No. 6,292,108, cited above. - The
sensor module 12 includes a microelectronic central processing unit (CPU) 21, also referred to more generally and more commonly as “a processor.” TheCPU 21 is interfaced to theaccelerometer 20 through a suitable I/O interface 22 according to specifications of theCPU 21 and theaccelerometer 20. TheCPU 21 is also connected by suitable buses to a radio transmitter/receiver circuit 23 of a type known in the art for converting between digital data and radio signals encoded with the data. The radio transmitter/receiver circuit 23 transmits and receives these signals through asuitable antenna 24. TheCPU 21 is operated by executing instructions in acontrol program 25 stored inmemory 26. Thememory 26 is shown generally and may include both memory on-board, and external memory in addition to that available on-board the CPU, and also includes memory for data storage in addition to the control program. Thesensor module 12 also has apower source 27, which may be a battery, a connection to an AC power outlet or a solar power transducer and storage circuit. Thesensor module 12 also includes a smallvisual display 28 and analarm 29 having an audible mode of operation and an inaudible, vibratory mode of operation. - Referring to
FIG. 3 , thehost computer system 17, which can be addressed at an Internet address, has anetwork server portion 30 for transmitting and receiving data on the Internet. The host computer system also has a microelectronic central processing unit (CPU) 31, also referred to more generally and more commonly as “a processor.” TheCPU 31 is operated by executing instructions in acontrol program 32 stored inmemory 33. Thememory 32 is shown generally and may include both memory on-board, and external memory in addition to that available on theCPU 31, and also includes memory for data storage in addition to the control program. Thehost computer system 17 includes avisual display 34 and also has the ability to emit audible sounds through a built-inspeaker 36. Thehost computer system 17 may also include adatabase 35, which can be stored in a computer memory or in an external bulk storage memory device of a type known in the art. - The methods disclosed herein can be used to continuously monitor civil structures or large manufacture items to detect impending failures, thereby detecting problems before they occur and thus allowing condition-based maintenance. The method is based on a premise that hysteresis strain energy (HSE) per unit time is the same as mechanical power invested in crack growth per unit time. Mechanical power can be computed from the force (a vector quantity), F, as mass times acceleration (a vector quantity), A, or F=mA. Also, velocity, V, (a vector quantity) is the time integral of acceleration, V=∫dt A. Mechanical power P (a scalar quantity) is the result of computing a vector dot-product of force and velocity, P=F·V. Consequently, mechanical (vibration) power in a complex structure can be determined by analysis of the tri-axial acceleration. Namely, the power per unit mass is P/m=F·V/m, so mechanical power can be computed from acceleration data as
-
P/m=A·∫dtA (1) - U.S. Pat. No. 6,460,012, assigned to the assignee herein, discloses a method for analysis of a load, L(t), and deflection, δ(t) versus time, t. The deflection can be expressed in relation to strain energy, U, by integration over the Nth load-unload cycle (∫C) of the structure as follows:
-
∫C Ldδ=dU/dN. (2) - The chain rule expansion of the right-hand side of Equation (2) in crack length, a, gives:
-
∫C Ldδ(dU/da)×(da/dN). (3) - The Griffith-energy criterion states that a crack cannot grow unless the strain energy per unit of crack length, dU/da, equals (or exceeds) the Griffith energy:
-
dU/da≦G 1C. (4) - G1C is a material property, and the subscript, 1, is for mode-one failure (under tension). Substitution of Equation (4) for the case of equality into Equation (3) yields:
-
∫C Ldδ=G 1c(da/dN). (5) - Consequently, the acquisition of load and deflection to obtain hysteresis strain energy, ∫C L dδ, allows inference of the crack growth rate by re-arranging Equation (5):
-
da/dN=∫ C L dδ/G 1C. (6) - Each load-unload cycle spans a cycle time, Δt, and corresponds to multiplication of the left-hand side of Eq. (5) by dN/dt, while the right-hand side is divided by Δt. The resultant form is then:
-
(da/dN)(dN/dt)=(1/Δt)∫C Ldδ/G 1C. (7) - The form, (da/dN)(dN/dt), is simply the crack growth rate, da/dt, from the chain rule. The HSE over the cycle time, (1/Δt)∫C L dδ, is the hysteresis power that goes into growing the crack at this rate, da/dt. Consequently, Eq. (7) can be rewritten as:
-
da/dt=P/G 1C. (8) - The mechanical vibration power from Eq. (1) can then be substituted in Eq. (8) to yield:
-
da/dt=A·∫dtA. (9) - Note that the mass (m) and Griffith energy (G1C) have been dropped from Eq. (8), because they simply scale the equation by constant factor (m/G1C). Eq. (9) means that the conversion of tri-axial acceleration (vector) to mechanical power (a scalar quantity) allows direct monitoring of the crack growth rate, as (for example) illustrated in
FIG. 5 . Because mechanical power (P) is a scalar quantity, this result is independent of the orientation of any tri-axial accelerometer in the case where acceleration is sensed along three axes which are orthogonal to each other. -
FIG. 5 shows that structural failure is indicated by a statistically significant change in the hysteresis strain energy (HSE), as an outlier in HSE slope beyond four standard deviations (4σ).FIG. 5 shows the slope of HSE versus number of fatigue cycles for corroded (10% thickness reduction), un-notched aluminum sample with typical failure forewarning in the last 1-5% of life. More specifically, thetop curve 61 inFIG. 5 is +4σ above the running value of the slope of HSE. Thebottom curve 62 inFIG. 5 is −4σ below the running value of the slope of HSE. Themiddle curve 63 is the HSE slope, which falls abruptly below the −4σ level at 49 000 cycles, as an indicator of structural failure. - This approach is also valid for mode-two (in-plane shear, or flexure) and mode-three (out-of-plane shear) failures. The method predicts failure within the last 1-20% of the total life cycles for a variety of materials (aluminum, steel, and glass-fiber/polymer matrix composite) across all three failure modes under a wide range of failure conditions: low-temperature creep crack growth; stress corrosion cracking; fatigue cracking; (un)corroded states; single- and multiple-site or widespread fatigue damage; constant amplitude loading, and periodic overloads. Further details are provided in Welch et al. in U.S. Pat. No. 6,460,012, which is hereby incorporated by reference.
-
FIG. 4 illustrates a method for providing forewarning of a critical event, such as a structural failure in a civil structure. This method is performed by one or more processors in executing instructions in one or more control programs, depending on the network configuration. In a first embodiment, the acceleration data is transmitted from thesensors 11 to thehost computer system 17 which performs the operations diagrammed inFIG. 4 . In a second embodiment, the operations are performed by each of thesensors 11 and a forewarning signal is signaled from thesensors 11 in the first instance, and may be relayed to the host computer. - This process of forewarning in other environments was disclosed in prior patents of one of the inventors herein, which are referred to in the Background of the Invention, including U.S. Pat. No. 7,209,861. In this application, the process shown in
FIG. 4 is carried out in a computing apparatus inFIGS. 2 and 3 , through execution of a program of computer instructions, represented by the blocks inFIG. 4 beginning with an initialization of parameters step represented bystart block 40. Next, as represented byprocess block 41, acceleration data for multiple axes are acquired, either through a network or from memory, and are converted to a process-indicative scalar signal for mechanical power, P, according to Eq. (9) above. The mechanical power data, P, is sampled at equal time intervals, τ, starting at an initial time, t0, yielding a sequence of N points, called a “cutset,” ei=e(t0+iτ). This is a subset of the dataset being analyzed. - Next, as represented by
process block 42, artifacts are removed from the data with a zero-phase quadratic filter that performs better than conventional filters. This filter uses a moving window of 2w+1 points of data, with the same number of data points, w, on either side of a central point. A parabola is fitted in the least-squares sense to these data points, and the central point is selected to estimate the low-frequency artifact, fi. The residual (artifact-filtered) signal, gi=ei−fi, has essentially no low-frequency artifact activity. All subsequent analysis uses this artifact-filtered data, gi. This is known in the art from U.S. Pat. No. 5,626,145. - Next, as represented by
decision block 43, a quality check is made of the entire dataset as described in U.S. Pat. No. 5,815,412, col. 12, line 64 to col. 14, line 9. If poor data quality is detected the dataset is discarded, as represented by the “FAIL” result branch and the process returns to obtain a new set of data atprocess block 41. If the quality is suitable as represented by the “PASS” result branch, the routine proceeds to thenext decision block 44. - A test is then made for whether this is the first cutset of data as represented by
decision block 44. If so, as represented by the “YES” result, the minimum, and maximum, gmax, in the data are calculated of the first baseline cutset as represented by process block and 45 inFIG. 4 . Each artifact-filtered point is converted into a discrete symbol, si, as one of S different integers in the range, 0≦si≦S−1. Contiguous, non-overlapping partitions are selected to obtain uniform symbols: si=INT[S (gi−gmin)/(gmax−gmin)] for gi<gmax, and si=S−1 for gi=gmax to maintain exactly S discrete symbols as represented byprocess block 46 inFIG. 4 . The function, INT, converts a decimal number to the next lowest integer [e.g., INT (3.14)=3]. - Alternatively, equiprobable symbols are formed by ordering all N of the base case artifact-filtered, time-serial data points from the smallest to largest value. The first N/S of these ordered values correspond to the first symbol, 0. Ordered data values (N/S)+1 through 2N/S correspond to the second symbol, 1, and so on. Equiprobable symbols have non-uniform partitions in the signal amplitude with the same occurrence frequency of gi values by construction, and thus have no information about the dynamical structure. In contrast, symbols with uniform partitions (uniform symbols), si=S (gi−gmin)/(gmax−gmin), have inherent dynamical structure before beginning the (PS) reconstruction, where gmax and gmin are the maximum and minimum values of the gi-data, respectively. Thus, one advantage of equiprobable symbols is that dynamical structure arises only from the PS reconstruction, as described below. Moreover, large negative and large positive values of gi have little effect on equiprobable symbolization, but dramatically change the partitions for uniform symbols.
- Next, as represented by
process block 47, the time-serial si-data are converted into a geometric object via phase-space (PS) reconstruction that uses time-delay vectors with the form, -
y(i)=[s i , s i+λ , . . . , s i+(d−1)λ], (10) - which partitions the PS into Sd hypercubes or bins. Each bin (or state) can be identified by a unique integer, J, via base-S arithmetic, namely, J=Σmsi+λSm, where the summation, Σm, is over the
range 0≦m≦d−1. Additional data channels may add more information about the data, implying that a multi-channel PS-vector could contain more information than a single channel. The multi-channel PS vector takes the form: -
y(i)=[s i(1), s i+λ(1), . . . , s i+(d−1)λ(1), . . . , s i(Ch.), s i+λ(Ch.), . . . , s i+(d−1)λ(Ch.)] (11) - Here, the symbol, s(k) denotes values from the k-th channel, i.e., 1≦k≦Ch., where Ch. is the total number of data channels. Now, the symbolization divides the multi-channel PS in SdCh. bins, where the bin identifier is J=Σk Σmsi+mλ(k)Sm+d(k−1). The choice of lag, λ, and embedding dimension, d, determines how well the PS reconstruction unfolds the dynamics. An excessively large embedding dimension could result in over-fitting of real data with finite length and noise.
- Conversion of the time-serial data into discrete PS states allows the construction of a statistical distribution function (DF), as represented by
process block 48, by counting the number of PS points that occur in each bin. The population of the ith bin of the DF, is denoted as Qi, for the base case, and Ri for a test case, respectively. An (un)changing DF indicates (un)altered dynamics. Next, a check is made to determine if the baseline number “B” of cutsets have been processed, as represented bydecision block 49. The dissimilarity functions, DFs, from first B cutsets are saved as represented byblocks 50 inFIG. 1 , to represent the nominal baseline dynamics. As represented bydecision block 51, when the number of cutsets equals the statistical sample number, B, the baseline dissimilarity calculations are completed as represented byprocess block 52. - The baseline DFs correspond to the nominal-state dynamics. These B baseline DFs are exhaustively compared to one another in pair-wise fashion. This comparison is via the dissimilarity measures (DM) of Eqs. (12)-(15), as represented by
block 54. V denotes the phase-space dissimilarity measures from the set, V={LN, LC, χN 2, χC 2}.V is the mean value of the dissimilarity measure, which is computed by comparison among the B(B−1)/2 unique combinations of the B base case cutsets, along with a corresponding standard deviation, σ1, for each DM from the set, V. The dissimilarity measures are computed as represented byblock 53, between DFs for the baseline, Qi, and test case, Ri, respectively. One set of dissimilarity measures (DM) for the non-connected phase space is: -
- The summations in Eqs. (12)-(13) run over all of the populated PS bins. The χ2 statistic is one of the most powerful, robust, and widely used tests for dissimilarity between two distribution functions (DFs). In this work, χ2 is not an unbiased statistic for accepting or rejecting a null statistical hypothesis but rather is a relative measure of dissimilarity between the two distribution functions (DFs). The L1 distance is the natural metric for the distribution functions (DFs) by its direct relation to the total invariant measure on the attractor. These measures account for changes in the geometry and visitation frequency of the attractor. Consistent calculation requires the same number of points in both the base and test case distribution functions (DFs), identically sampled; otherwise the distribution functions must be resealed.
- The accuracy and sensitivity of the PS reconstruction can be enhanced by connecting successive PS points as prescribed by the underlying dynamics, y(i)→y(i+μ), for μ≧1. Thus, we obtain a discrete representation of the process flow Y(i)=[y(i), y(i+μ)] that is formed by adjoining two successive vectors from the d-dimensional reconstructed PS. Y(i) is a 2d-dimensional, connected-phase-space (CPS) vector. As before, Q and R denote the CPS DFs for the base case and test case, respectively. We then define the measures of dissimilarity between these two CPS DFs via the L1-distance and χ2 statistic, as before:
-
- The subscript c denotes connected phase space measures in Eqs. (14)-(15), while the subscript, N, in Eqs. (12)-(13) denotes non-connected phase space measures. The subscripts, J and K, are identifiers for the two successive PS states, y(i) and y(i+μ), respectively. Connected phase space measures have higher discriminating power than their non-connected counterparts. The quantities in Eqs. (11)-(14) are referred to as phase space dissimilarity measures (PSDM). Their definitions allow PSDM to detect transitions between regular and chaotic regimes, as well as to discriminate between different chaotic regimes. While straightforward methods exist for discriminating between regular and chaotic motion, or for detecting the transition between these regimes, discriminating between close chaotic regimes via traditional nonlinear measures (e.g., Lyapunov exponents, Kolmogorov entropy, and correlation dimension) is almost impossible. The reason for the superior performance of PSDM is rather simple: traditional nonlinear measures use a difference of averages, while PSDM use sums of the absolute value of differences.
- Next, the dissimilarity measures are normalized by comparing each of the B baseline cutsets to each (ith) test case cutset, and then computing the corresponding average dissimilarity value, Vi, of the ith cutset as represented by
process block 55. The normalized form is: U(V)=|Vi−V |/σ1, which is the number of standard deviations that the test case deviates from the baseline mean. Each contiguous, non-overlapping test case cutset is compared to each of the B base case cutsets, to obtain the corresponding average dissimilarity value, Vi, of the i-th analysis window for each dissimilarity measure. A statistically significant trend in the normalized dissimilarity measure indicates equipment degradation for failure forewarning. - Once the normalized measures for the test and base cases have been obtained, a forewarning status can be determined, as represented by
process block 56, by computing a composite measure of dissimilarity (Ci=LN+LC+χN 2+χC 2) from the normalized measures of dissimilarity (U). A forewarning threshold is then developed by computing a cumulative sum of the composite measure of dissimilarity (Σi Ci) over a series of time frames. A straight-line fit is applied to the cumulative sum of composite measures (Σi Ci) using a least squares calculation. Then, a standard deviation (σ2) is computed between the straight-line fit and the cumulative sum (Σi Ci). Event forewarning corresponds to the difference between the cumulative sum (Σi Ci) and the straight line fit exceeding a forewarning threshold by more than a selected number of standard deviations. In response to the composite measure (Ci) exceeding a forewarning threshold for a preselected number of occurrences, as represented by the “Yes” result fromdecision block 57, a forewarning signal of an impending critical event is provided to a human observer, as represented by I/O block 58, through one of or a combination of a visual signal, an audible signal or a tangible signal. An end-of-life failure forewarning threshold can also be determined and an end-of-life failure forewarning signal issued based on a number of successive occurrences of Ci above a failure threshold over a selected time period, with the failure threshold being computed as disclosed in U.S. Pat. No. 7,209,861, col. 10,line 25 to col. 11,line 26. In an extension of this method disclosed in copending U.S. patent application Ser. No. 12/872,471, which is assigned to the assignee herein, a failure forewarning threshold for detecting failure forewarnings corresponds to a minimum in said standard deviation (σ2) in relation to said straight-line fit of Σi Ci, which is the cumulative sum of the composite measure of dissimilarity over a sampling period. A failure value threshold is determined based on a number of failure forewarnings occurring after this minimum in the standard deviation is detected. In this case a forewarning of the impending critical event is signaled to a human observer, as represented inblock 58, after either one of 1) a selected number of occurrences of the composite measure of dissimilarity (Ci) above the forewarning threshold or 2) after a selected number of occurrences of the composite measure of dissimilarity (Ci) above the failure threshold. - As represented by
decision block 59, if the result from executingdecision block 57 is “No,” a check is then made for more data, and if more data is available, the routine returns to block 41, to continue the analysis for the next dataset via the above algorithm, and then combine the results for all of the datasets. If no more data is available, the execution is ended as represented byblock 60. - This method can be applied to any large physical structure (e.g., bridge, aging aircraft, cranes, spacecraft, buildings) to forewarn of an impending structural failure under arbitrary loading, thus reducing the cost of inspections and repairs and conserving critical infrastructure. The method to perform this analysis has been demonstrated on desktop computers. Modern hand-held devices (e.g., PalmPRE and iPhone) are capable (in terms of CPU speed, memory, and on-board data storage) of carrying out the method as well. Such hand-held devices have embedded GPS (to provide geo-location), wireless communication (to send the forewarning information), and a tri-axial accelerometer (to acquire the acceleration data). Therefore, these devices could also be re-packaged and utilized as the
sensors 11. And, a cellular network can be used in place of the Internet wide area network disclosed herein between thesensors 11 and a host computer in a cellular network. - After executing the program of instructions according to the methods described above, and as seen in
FIGS. 2 and 3 , theCPU 21 or theCPU 31 can transmit a signal through an output interface circuit of a type well known in the art to alarmdevices screen display - This has been a description of examples of the invention. It will apparent to those of ordinary skill in the art that certain modifications might be made without departing from the scope of the invention, which is defined by the following claims.
Claims (25)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/962,787 US8410952B2 (en) | 2010-12-08 | 2010-12-08 | Methods for forewarning of critical condition changes in monitoring civil structures |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/962,787 US8410952B2 (en) | 2010-12-08 | 2010-12-08 | Methods for forewarning of critical condition changes in monitoring civil structures |
Publications (2)
Publication Number | Publication Date |
---|---|
US20120146802A1 true US20120146802A1 (en) | 2012-06-14 |
US8410952B2 US8410952B2 (en) | 2013-04-02 |
Family
ID=46198792
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/962,787 Active 2031-09-15 US8410952B2 (en) | 2010-12-08 | 2010-12-08 | Methods for forewarning of critical condition changes in monitoring civil structures |
Country Status (1)
Country | Link |
---|---|
US (1) | US8410952B2 (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110304499A1 (en) * | 2008-10-15 | 2011-12-15 | Fraunhofer-Gesellschaft Zur Forderung Der Angewandten Forschung E.V. | System for determining the movement of a swaying structure |
US8410952B2 (en) * | 2010-12-08 | 2013-04-02 | Ut-Battelle, Llc | Methods for forewarning of critical condition changes in monitoring civil structures |
WO2013175162A1 (en) * | 2012-05-22 | 2013-11-28 | Aes Engineering Ltd | A monitoring device |
US20130342343A1 (en) * | 2012-06-22 | 2013-12-26 | K.L. Harring Transportation LLC | Cargo tracking and monitoring system |
WO2014142701A1 (en) * | 2013-03-14 | 2014-09-18 | Siemens Aktiengesellschaft | A system and method for evaluating a dam state |
US20150088428A1 (en) * | 2013-09-23 | 2015-03-26 | San Diego State University Research Foundation | Computationally efficient nonlinear structural analysis |
FR3036499A1 (en) * | 2015-05-22 | 2016-11-25 | Explorcam | ALARM DEVICE |
JP6429215B1 (en) * | 2017-12-12 | 2018-11-28 | 株式会社構研エンジニアリング | Inspection method for structures |
ES2697689A1 (en) * | 2017-07-24 | 2019-01-25 | Idvia 2020 Horizonte 2020 S L | DEVICE AND PROCEDURE FOR MONITORING SOCAVATION (Machine-translation by Google Translate, not legally binding) |
CN110659829A (en) * | 2019-09-24 | 2020-01-07 | 北京理工大学 | Autonomous spacecraft unexecuted planning sequence resource abnormity detection and repair method |
EP3242972B1 (en) | 2015-01-05 | 2020-07-15 | BAE Systems PLC | Mobile bridge apparatus |
CN114062421A (en) * | 2020-07-31 | 2022-02-18 | 麦克汰股份有限公司 | Method and system for diagnosing life of structure |
US20220117039A1 (en) * | 2020-10-12 | 2022-04-14 | Sercel | System and method for generating and collecting vibration data with a view to monitoring a structure |
US20220303747A1 (en) * | 2021-03-16 | 2022-09-22 | Omidreza Ghanadiof | Earthquake Damage Warning System |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TR201620208A1 (en) * | 2016-12-30 | 2018-07-23 | Ford Otomotiv Sanayi As | COMPACT VIBRATION AND NOISE MAPPING SYSTEM AND METHOD |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6292108B1 (en) * | 1997-09-04 | 2001-09-18 | The Board Of Trustees Of The Leland Standford Junior University | Modular, wireless damage monitoring system for structures |
US6807862B2 (en) * | 2002-02-21 | 2004-10-26 | Sekos, Inc. | Device and method for determining and detecting the onset of structural collapse |
US20110137586A1 (en) * | 2009-07-06 | 2011-06-09 | Yong Jiang | Wind turbine monitoring and adjusting |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU1820697A (en) | 1995-12-15 | 1997-07-17 | Structural Integrity Monitoring Systems, Inc. | Structural monitoring sensor system |
US5815413A (en) | 1997-05-08 | 1998-09-29 | Lockheed Martin Energy Research Corporation | Integrated method for chaotic time series analysis |
US6006163A (en) | 1997-09-15 | 1999-12-21 | Mcdonnell Douglas Corporation | Active damage interrogation method for structural health monitoring |
US6460012B1 (en) | 1999-09-16 | 2002-10-01 | U.T. Battelle, Llc, | Nonlinear structural crack growth monitoring |
US6459050B1 (en) | 1999-09-20 | 2002-10-01 | Ut-Battelle, Inc. | Method and appartus for converting static in-ground vehicle scales into weigh-in-motion systems |
US6484132B1 (en) | 2000-03-07 | 2002-11-19 | Lockheed Martin Energy Research Corporation | Condition assessment of nonlinear processes |
US8384538B2 (en) | 2002-06-11 | 2013-02-26 | Intelligent Technologies International, Inc. | Remote monitoring of fixed structures |
US7139677B2 (en) | 2002-07-12 | 2006-11-21 | Ut-Battelle, Llc | Methods for consistent forewarning of critical events across multiple data channels |
US7209861B2 (en) | 2002-07-12 | 2007-04-24 | Ut-Battelle Llc | Methods for improved forewarning of critical events across multiple data channels |
US20040078170A1 (en) | 2002-10-17 | 2004-04-22 | Don Di Marzio | System and method for monitoring a structure |
JP2006509551A (en) | 2002-12-10 | 2006-03-23 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | Activity monitoring |
US20050037730A1 (en) | 2003-08-12 | 2005-02-17 | Albert Montague | Mobile wireless phone with impact sensor, detects vehicle accidents/thefts, transmits medical exigency-automatically notifies authorities |
US7561200B2 (en) | 2004-07-26 | 2009-07-14 | Csi Technology, Inc. | Apparatus and method for automation of imaging and dynamic signal analyses |
JP4592360B2 (en) | 2004-09-02 | 2010-12-01 | 公立大学法人会津大学 | Physical condition monitoring device |
US7451642B2 (en) | 2004-09-14 | 2008-11-18 | Bridgestone Corporation | System and method for quantitative analysis of cause of tire trouble |
US20060248954A1 (en) | 2005-04-26 | 2006-11-09 | Snieder Roelof K | System for and method of monitoring structural integrity of a structure |
US8024980B2 (en) | 2008-01-24 | 2011-09-27 | Microstrain, Inc. | Independently calibrated wireless structural load sensor |
WO2010034022A1 (en) | 2008-09-22 | 2010-03-25 | Purdue Research Foundation | Methods and apparatus for diagnosing faults of a vehicle |
US8410952B2 (en) * | 2010-12-08 | 2013-04-02 | Ut-Battelle, Llc | Methods for forewarning of critical condition changes in monitoring civil structures |
-
2010
- 2010-12-08 US US12/962,787 patent/US8410952B2/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6292108B1 (en) * | 1997-09-04 | 2001-09-18 | The Board Of Trustees Of The Leland Standford Junior University | Modular, wireless damage monitoring system for structures |
US6807862B2 (en) * | 2002-02-21 | 2004-10-26 | Sekos, Inc. | Device and method for determining and detecting the onset of structural collapse |
US20050125197A1 (en) * | 2002-02-21 | 2005-06-09 | Ziyad Duron | Device and method for determining and detecting the onset of structural collapse |
US7228240B2 (en) * | 2002-02-21 | 2007-06-05 | Ziyad Duron | Device and method for determining and detecting the onset of structural collapse |
US20110137586A1 (en) * | 2009-07-06 | 2011-06-09 | Yong Jiang | Wind turbine monitoring and adjusting |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8604964B2 (en) * | 2008-10-15 | 2013-12-10 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | System for determining the movement of a swaying structure |
US20110304499A1 (en) * | 2008-10-15 | 2011-12-15 | Fraunhofer-Gesellschaft Zur Forderung Der Angewandten Forschung E.V. | System for determining the movement of a swaying structure |
US8410952B2 (en) * | 2010-12-08 | 2013-04-02 | Ut-Battelle, Llc | Methods for forewarning of critical condition changes in monitoring civil structures |
GB2516998A (en) * | 2012-05-22 | 2015-02-11 | Aes Eng Ltd | A monitoring device |
WO2013175162A1 (en) * | 2012-05-22 | 2013-11-28 | Aes Engineering Ltd | A monitoring device |
US20130342343A1 (en) * | 2012-06-22 | 2013-12-26 | K.L. Harring Transportation LLC | Cargo tracking and monitoring system |
US9501920B2 (en) * | 2012-06-22 | 2016-11-22 | K.L. Harring Transportation LLC | Cargo tracking and monitoring system |
WO2014142701A1 (en) * | 2013-03-14 | 2014-09-18 | Siemens Aktiengesellschaft | A system and method for evaluating a dam state |
US20150088428A1 (en) * | 2013-09-23 | 2015-03-26 | San Diego State University Research Foundation | Computationally efficient nonlinear structural analysis |
EP3242972B1 (en) | 2015-01-05 | 2020-07-15 | BAE Systems PLC | Mobile bridge apparatus |
FR3036499A1 (en) * | 2015-05-22 | 2016-11-25 | Explorcam | ALARM DEVICE |
ES2697689A1 (en) * | 2017-07-24 | 2019-01-25 | Idvia 2020 Horizonte 2020 S L | DEVICE AND PROCEDURE FOR MONITORING SOCAVATION (Machine-translation by Google Translate, not legally binding) |
WO2019020851A1 (en) * | 2017-07-24 | 2019-01-31 | Idvia 2020 Horizonte 2020, S.L. | Device and method for scour monitoring |
JP6429215B1 (en) * | 2017-12-12 | 2018-11-28 | 株式会社構研エンジニアリング | Inspection method for structures |
CN110659829A (en) * | 2019-09-24 | 2020-01-07 | 北京理工大学 | Autonomous spacecraft unexecuted planning sequence resource abnormity detection and repair method |
CN114062421A (en) * | 2020-07-31 | 2022-02-18 | 麦克汰股份有限公司 | Method and system for diagnosing life of structure |
US20220117039A1 (en) * | 2020-10-12 | 2022-04-14 | Sercel | System and method for generating and collecting vibration data with a view to monitoring a structure |
US11743972B2 (en) * | 2020-10-12 | 2023-08-29 | Sercel | System and method for generating and collecting vibration data with a view to monitoring a structure |
US20220303747A1 (en) * | 2021-03-16 | 2022-09-22 | Omidreza Ghanadiof | Earthquake Damage Warning System |
US11477631B2 (en) * | 2021-03-16 | 2022-10-18 | Omidreza Ghanadiof | Earthquake damage warning system |
Also Published As
Publication number | Publication date |
---|---|
US8410952B2 (en) | 2013-04-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8410952B2 (en) | Methods for forewarning of critical condition changes in monitoring civil structures | |
He et al. | Integrated structural health monitoring in bridge engineering | |
Weng et al. | Output-only modal identification of a cable-stayed bridge using wireless monitoring systems | |
Nair et al. | Time series-based damage detection and localization algorithm with application to the ASCE benchmark structure | |
TWI464443B (en) | Artificial intelligence earthquake early warning method and earthquake detecting system | |
Kiremidjian et al. | Structural damage monitoring for civil structures | |
WO2020188585A1 (en) | System and method for structural health monitoring using internet of things and machine learning | |
CN110780347B (en) | Earthquake destructive power prediction device and method based on cyclic neural network | |
CN107807381B (en) | Dynamic monitoring method and device for slope instability risk based on rock mass fracture microseism wave activity rule | |
JP2008003043A (en) | Method, apparatus, and system for detecting damages | |
CN115022187B (en) | Situation awareness method and device for electric-gas comprehensive energy system | |
CN108520227A (en) | A kind of Bridge Structural Damage localization method of the transfer entropy based on dual sensor information | |
US8416086B2 (en) | Methods for improved forewarning of condition changes in monitoring physical processes | |
Moeinifard et al. | Lost vibration test data recovery using convolutional neural network: a case study | |
Diao et al. | Wavelet entropy based structural damage identification under seismic excitation | |
CN113109587A (en) | Method, system and device for measuring wind speed of power transmission line and storage medium | |
CN117189239A (en) | Tunnel surrounding rock damage monitoring method | |
Choi et al. | Efficient calculation of statistical moments for structural health monitoring | |
JP2003214849A (en) | Disaster-prevention monitoring system | |
EP4242582A1 (en) | Device and method for calculating behavior of linear structure | |
Łęczycki et al. | Extended sensor reliability evaluation method in multi-sensor control systems | |
CN112816122A (en) | Bolt tightness degree monitoring method based on deep learning and piezoelectric active sensing | |
Cetinkaya et al. | Deep Learning and Internet of Things (IoT) Based Monitoring System for Miners | |
CN110631634A (en) | Structure monitoring and evaluating method and system for complete inverse analysis and storage medium | |
CN117272697B (en) | Multi-sample-based method for calculating reliable service life of steel wire rope for ocean engineering equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: UT-BATTELLE, LLC, TENNESSEE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:ABERCROMBIE, ROBERT K.;HIVELY, LEE M.;REEL/FRAME:025757/0699 Effective date: 20110207 |
|
AS | Assignment |
Owner name: U.S. DEPARTMENT OF ENERGY, DISTRICT OF COLUMBIA Free format text: CONFIRMATORY LICENSE;ASSIGNOR:UT-BATTELLE, LLC;REEL/FRAME:025935/0074 Effective date: 20110126 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YR, SMALL ENTITY (ORIGINAL EVENT CODE: M2552); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY Year of fee payment: 8 |